Objective color harmony assessment for visible and infrared color fusion images of typical scenes
| Autor: | Ling-Xue Wang, Gao Shaoshu, Weiqi Jin |
|---|---|
| Rok vydání: | 2012 |
| Předmět: |
Image fusion
Color histogram Harmony (color) Color quality business.industry Color image Computer science Color normalization ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION General Engineering Colorfulness Color balance False color Image segmentation Color space Atomic and Molecular Physics and Optics Color quantization Color model Computer vision Artificial intelligence business ComputingMethodologies_COMPUTERGRAPHICS |
| Zdroj: | Optical Engineering. 51:117004 |
| ISSN: | 0091-3286 |
| DOI: | 10.1117/1.oe.51.11.117004 |
| Popis: | For visible and infrared color fusion images of three typical scenes, color harmony computational models are proposed to evaluate the color quality of fusion images without reference images. The models are established based on the color-combination harmony model and focus on the influence of the color characteristics of typical scenes and the color region sizes in the fusion image. For the influence of the color characteristics of typical scenes, color harmony adjusting factors for natural scene images (green plants, sea, and sky) are defined by measuring the similarity between image colors and corresponding memory colors, and that for town and building images are presented based on the optimum colorfulness range suited for a human. Simultaneously, considering the influence of color region sizes, the weight coefficients are established using areas of the color regions to optimize the color harmony model. Experimental results show that the proposed harmony models are consistent with human perception and that they are suitable to evaluate the color harmony for color fusion images of typical scenes. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|